Sequential, multiple assignment, randomized trials (SMARTs) allow investigation of experimental treatment regimens in which individuals are successively randomized to different treatments based on their responsiveness to previous treatment. Successful implementation depends on identifying patients who respond to treatment, though in some situations such mechanisms may not exist. We propose a method for probabilistically assigning responder statuses to subjects completing the first stage of a SMART-like trial supplemented with information provided through pilot data. Two approaches for estimating and updating these probabilities are discussed, including a combination of cluster analysis and discriminant analysis as well as a mixture-model approach. In both cases the estimated responder probabilities are used to adaptively allocate patients between responder and non-responder classifications, allowing patients to continue onto the second phase of treatment. Simulation studies have shown that both methods have acceptable misclassification error rates, bias, and efficiency although the mixture-model approach performs better and with fewer computational challenges.